1. Field of the Invention
This invention relates to electric- or fluid-powered torque-applying tools, and more specifically, to control methods for such power tools.
2. Description of Related Art
Computer-controlled fluid- or electric-power tools are typically used in production environments to secure threaded fasteners (e.g., nuts and bolts) into joints. Such power tools typically include a handheld unit coupled to a controller. The handheld unit, or tool, usually has a high-speed, high-torque motor coupled to a universal adapter head. Various interchangeable bits are connected to the head in order to drive threaded fasteners, e.g., bits appropriate for hex-head bolts and hexagonal nuts. The motor of each handheld unit is usually rated to apply no more than a maximum amount of torque, and is also usually rated to run at no more than a maximum speed.
The controller for each handheld unit controls the power supply for each handheld unit, and also monitors such parameters as the current tool speed and current applied torque. In a typical fastening job, fasteners are tightened to a predetermined, specified torque. As the handheld units operate at high speed, on the order of several hundred RPM, the controller is typically used to start and stop the motor of the handheld unit automatically so that the torque applied to the fastener and joint does not exceed the specified torque or the torque rating of the tool""s motor.
The high speed at which the tool""s motor operates means that a single fastening job, for example tightening a single bolt, may only require a few milliseconds. Therefore, even though the tools are computer-controlled, there is a substantial likelihood that the tool will xe2x80x9covershootxe2x80x9d the desired torque, thus increasing the stress on the joint and potentially damaging the tool.
Joints are usually classified by their torque/turn rates. The torque/turn rate is defined as the ratio of change in torque per unit of rotation of the fastener. A fastening application is considered xe2x80x9csoftxe2x80x9d when the torque/turn rate is low, and is considered xe2x80x9chardxe2x80x9d when the torque/turn rate is high. Joints may also be classified as xe2x80x9cmediumxe2x80x9d or may be completely irregular in their properties.
In evaluating the characteristics of a fastening job, the tool""s torque/time rate is also important. The torque/time rate is dependent upon the rotational speed of the tool and the torque/turn rate of the fastener itself, therefore, the torque/time rate is strongly influenced by the torque/turn rate. In general, a high torque/time rate is indicative of a xe2x80x9chardxe2x80x9d joint, while a relatively low torque/time rate is indicative of a xe2x80x9csoftxe2x80x9d joint. A high torque/time rate is one of the factors which contributes to the problem of torque xe2x80x9covershoot.xe2x80x9d
To remedy the problem of torque overshoot, the user may simply choose to run the tool at a lower rotational speed. Unfortunately, that simple solution is not practical in production environments because a slower-running tool takes more time to finish a tightening process, thereby decreasing worker productivity and potentially causing the tool motor to overheat.
Because of the variability in joint properties, it becomes difficult to design a computer algorithm to properly control a torque-applying tool. Previous attempts have resulted in algorithms with somewhat limited utility.
Commonly-assigned U.S. Pat. No. 5,315,501 to Whitehouse discloses an algorithm for controlling power tools. The algorithm determines an internal torque target based upon the torque/turn or torque/time rate of the joint and dynamic characteristics of the tool, where the rate is calculated based on the controller-observed properties of several different joints. For optimum results, this algorithm must be used on a joint in which the amount of torque monotonically increases. This type of joint is not often encountered in practice, thus limiting the applicability of the algorithm. In cases where the algorithm can be applied, the controller is required to analyze up to 75% of the torque/turn characteristic of the joint before enabling the control algorithm. Given that xe2x80x9chardxe2x80x9d joints can be tightened in less than 10 ms, the controller usually does not have sufficient time to shut the tool down when the internal torque target is reached, resulting in torque overshoot. The method also preferably requires the use of an angular measuring device, which increases the cost of the system. Moreover, the angular measurements required for the method may not be accurate, because in a typical tool, the angular measurement changes if the user changes the position of the tool while tightening a joint.
Commonly-assigned U.S. Pat. No. 5,637,968 to Kainec et al. discloses an alternate method of power tool control in which the controller measures the torque/turn rate of the joint between 25% and 50% of the programmed target torque to classify the joint as either xe2x80x9csoft,xe2x80x9d xe2x80x9cmedium,xe2x80x9d or xe2x80x9chard.xe2x80x9d The controller then issues a command to execute an immediate downshift in speed based on the joint classification. The controller continues the tightening process at the reduced speed until the programmed target torque is reached.
The design disclosed in the Kainec patent can be improved in several ways. First, on a xe2x80x9chardxe2x80x9d joint, if the motor speed is reduced at 50% of the target torque and the tool is running at, for example, 1500 RPM, the tool has only about 1.5 ms to slow down before the target torque is reached, whereas the typical response time for a control system can be much greater than that. The immediate downshift imposed by the Kainec method is undesirable because the immediate change in speed can induce damaging dynamic loads on the motor and gearing. Immediate downshifts also consume more power, because the controller attempts to dynamically brake the tool""s motor. Moreover, the dynamic braking process itself causes the tool and controller to heat up unnecessarily.
The classification system imposed by the Kainec method also imposes some limitations. By categorizing all joints into one of only three categories and requiring a specific, fixed downshift in tool speed for each category, the method may prevent the controller from tightening each joint optimally. For example, a joint with a characteristic between xe2x80x9chardxe2x80x9d and xe2x80x9csoftxe2x80x9d would be classified as a xe2x80x9cmediumxe2x80x9d joint, and the controller would reduce speed at 50% of the target torque, which would actually increase the amount of time it takes to fasten the joint. For most joints, the Kainec method actually increases the amount of time it takes to fasten the joint. Additionally, the method does not account for the tool""s speed, so even when a joint is correctly classified, torque overshoot may still be a problem because a faster tool on a particular joint may require a greater reduction in tool speed in order to avoid overshoot.
Immediate downshifts in tool speed are also generally undesirable because of the manner in which torque-applying tools are tested. Typically, torque-applying tools are tested by using brakes to simulate the effects of tightening a threaded fastener. However, brakes have very high polar moments of inertia when compared to typical joint assemblies, which can affect the dynamic response time of the control system and an immediate downshift may cause instability, erratic torque readings, and thus, inaccurate test results.
Other power tool control algorithms that are commonly used include learning-based algorithms. A learning-based algorithm requires that the tool and controller be used on several xe2x80x9ctestxe2x80x9d joints of a particular type so that the controller can adapt the tool""s speed and performance to the characteristics of that particular joint. The controller runs the tool at various speeds until an optimum speed profile is determined for the particular type of joint. Unfortunately, once the learning algorithm is employed, the tool and controller may only be used optimally on the particular type of joint for which the controller has xe2x80x9ctrained.xe2x80x9d
One example of this type of learning-based algorithm is disclosed U.S. Pat. No. 5,215,270 to Udocon et al. The disclosed method applies continuous feedback control over the tool""s speed to maintain an xe2x80x9coptimumxe2x80x9d speed. The optimum speed is calculated by using an equation which includes empirical constants that must be determined for each joint. The method may either increase or decrease the tool""s speed to meet the calculated optimum speed.
Another example of a learning-based algorithm is disclosed in U.S. Pat. No. 5,650,574 to Sato et al. In this method, a family of tool speed profiles are precalculated and are stored in the tool controller. Over the course of several learning cycles, the controller implements each one of the family of tool speed profiles in succession, changing speed profile until the chosen tool speed profile causes no torque overshoot.
Aside from learning-based methods, some power tool control methods have employed continuous feedback control of tool speed based on the tool""s measured, applied torque, but these methods generally do not prevent torque overshoot. For example, U.S. Pat. No. 5,519,614 to Hansson discloses a similar continuous feedback control method in which the tool is held to a calculated, optimum rate of torque application, but the control method is used only to control the reaction forces experienced by the user.
There exists a need for an adaptive control method for a torque-applying tool that correctly fastens a joint to a specified torque without overshoot, regardless of the torque rate or class of the joint, or that does not require the torque-applying tool to test or learn a particular type of joint.
An exemplary method of controlling a torque-applying tool to apply a selected torque by controlling a speed of the torque applying tool includes calculating a first torque at an end of a deceleration ramp that is a percentage of the selected torque, calculating a second torque at a start of the deceleration ramp that is a percentage of the selected torque, calculating a first speed at the end of the deceleration ramp that is a percentage of a selected speed, periodically determining a peak torque applied by the tool, determining if the peak torque is greater than the first torque, stopping the tool, if the peak torque is greater than or equal to the first torque and the selected torque, determining if the peak torque is greater than or equal to the second torque, if the peak torque is not greater than the first torque, and calculating parameters descriptive of the deceleration ramp and controlling the speed in accordance with the parameters, if the peak torque is greater than or equal to the second torque.
Another exemplary method of controlling a torque applying tool to apply a selected torque by controlling a speed of the torque applying tool includes determining a final torque value, determining a final speed value, decreasing speed levels until the torque applying tool reaches the final torque value, periodically measuring the torque level until the final torque value is measured, and stopping the torque applying tool when the selected torque value is reached.
An exemplary torque-applying tool according to the invention includes a motor, a drive head that is driven by the motor, a sensor package including a torque sensor and a speed sensor, and a controller that calculates a first torque at an end of a deceleration ramp that is a percentage of a selected torque, calculates a second torque at a start of the deceleration ramp that is a percentage of the selected torque, calculates a first speed at the end of the deceleration ramp that is a percentage of a selected speed, periodically determines a peak torque applied by the tool, determines if the peak torque is greater than the first torque, stops the tool, if the peak torque is greater than or equal to the first torque and the selected torque, determines if the peak torque is greater than or equal to the second torque, if the peak torque is not greater than the first torque, and calculates parameters descriptive of the deceleration ramp and controls the speed in accordance with the parameters, if the peak torque is greater than or equal to the second torque.
An exemplary controller for a torque-applying tool including a motor, a drive head that is driven by the motor, a sensor package including a torque sensor and a speed sensor, calculates a first torque at an end of a deceleration ramp that is a percentage of a selected torque, calculates a second torque at a start of the deceleration ramp that is a percentage of the selected torque, calculates a first speed at the end of the deceleration ramp that is a percentage of a selected speed, periodically determines a peak torque applied by the tool determines if the peak torque is greater than the first torque, stops the tool, if the peak torque is greater than or equal to the first torque and the selected torque, determines if the peak torque is greater than or equal to the second torque, if the peak torque is not greater than the first torque, and calculates parameters descriptive of the deceleration ramp and controls the speed in accordance with the parameters, if the peak torque is greater than or equal to the second torque.